How to Train Your Own AI Model (LoRA) Using Personal or Favorite Celebrity Photos Without any GPU.

Akalanka Ekanayake
26 Sept 202303:04

TLDRIn this informative tutorial, viewers are guided through the process of training their own AI model using the Sea Art platform, specifically focusing on generating images with personalized or favorite celebrity faces. The video demonstrates how to create a dataset with images of Selena Gomez and set a trigger word for the model. It also covers the tagging process, model training with a base model like 'Dream Shaper,' and the generation of AI images using the trained model. The summary highlights the ease of use and the potential for creating unique AI-generated images, encouraging viewers to try the platform and share their creations.


  • 🎨 Start by visiting Sea Art's website and look for the 'Model and Training' option in the menu.
  • 📂 Click 'Create' to start setting up your dataset, providing a name and selecting a category.
  • 🖼️ Edit your dataset by uploading personal or favorite celebrity photos using drag-and-drop or the plus icon.
  • 🏷️ Use the 'Tag Settings' to set a model trigger word, such as the name of the celebrity.
  • 💡 After uploading, the platform suggests keywords for your images, which can be a helpful feature.
  • 👌 Ensure you are satisfied with the tags before saving your settings.
  • 🔧 There's an 'Advanced Config' section for fine-tuning, but beginners can stick with default settings.
  • 🚀 Initiate the training process by clicking 'Start Training' and providing a unique name for your model.
  • ☕ While training, which may take some time, you can relax and wait for the process to complete.
  • 📢 Upon completion, you'll receive a notification and can proceed to the 'LoRA' tab to select your model.
  • 🔑 To generate images, type in a prompt including your model's trigger word.
  • 🌟 Enjoy your first AI-generated image with the selected celebrity and experiment with different models for varied results.
  • 📝 Share your creations and experiences in the comments section to engage with the community.

Q & A

  • What is the main topic of the video transcript?

    -The main topic of the video transcript is training your own AI model using personal or favorite celebrity photos without the need for a GPU.

  • Which platform is used for AI image generation in the transcript?

    -The platform used for AI image generation in the transcript is 'sea art'.

  • How does one begin the process of creating a dataset on sea Art's website?

    -To begin creating a dataset on sea Art's website, one should look for the 'model and training' option in the menu, click on 'create,' and provide a name for the dataset along with selecting the appropriate category.

  • What is the purpose of the 'tag settings' option?

    -The 'tag settings' option is used to set model trigger words, which are keywords that will prompt the AI to generate images based on the trained model.

  • What does the platform suggest after uploading images?

    -After uploading images, the platform suggests keywords for your images, which can help in refining the AI model's understanding and generation of images.

  • What is the 'advanced config' section for?

    -The 'advanced config' section is for users who wish to fine-tune their model with more specific settings beyond the default options provided for beginners.

  • What is the next step after setting up the tags and configurations?

    -The next step is to click on 'start training,' provide a unique name for your model, and choose a base model such as 'dream shaper' to begin the training process.

  • How long does the training process typically take?

    -The training process might take a while, so the transcript suggests grabbing a coffee and relaxing while the model trains.

  • Once the model is trained, what happens next?

    -After the model is trained, you will receive a notification. You can then head to the 'LoRA' tab, choose your freshly trained model, and start generating images by clicking on 'generate' and typing in a prompt that includes your model's trigger word.

  • How can one switch between different models to generate images?

    -To switch between different models, one can go to the 'LoRA' tab and select the desired model before generating new images.

  • What is the final step suggested for users after generating their AI images?

    -The final step suggested is to share the creations in the comments section below the tutorial and to continue exploring and experimenting with the AI image generation feature.

  • What is the significance of including the model's trigger word in the prompt?

    -Including the model's trigger word in the prompt ensures that the AI generates images that are relevant to the trained model, which in this case is associated with the selected celebrity or personal photos.



🎨 Introduction to AI Image Generation with Sea Art

The video begins with an introduction to the concept of AI image generation, specifically mentioning the possibility of creating personalized images using one's own face or a favorite celebrity's face. The platform of choice for this tutorial is Sea Art, which offers a model training feature. The process starts by visiting the Sea Art website and selecting the 'model and training' option from the menu. The viewer is then guided to create a new dataset, which involves naming the dataset and choosing a category. The core of the demonstration involves using photos of Selena Gomez to illustrate how to add images to the dataset, either by dragging and dropping or using the plus icon. The video also covers setting model trigger words, which in this case is 'Selena,' and the platform's ability to suggest keywords for the images. After tagging, the user is instructed to save the settings and, if needed, explore the advanced configuration options for fine-tuning the model. The tutorial then moves on to the training process, where the user is prompted to start training, provide a unique model name, and select a base model like 'dream shaper.' The training process may take some time, so viewers are advised to take a break. Upon completion, the user receives a notification and can then proceed to the 'Laura' tab to select the trained model for image generation. The final step is to generate an image by typing in a prompt that includes the model's trigger word.



💡AI image generation

AI image generation refers to the process of creating images using artificial intelligence algorithms. In the context of the video, it involves using AI to generate images featuring the user's face or a favorite celebrity, making personalized images a reality.

💡Sea Art platform

The Sea Art platform is mentioned as the tool used for model training in the video. It is an online service that allows users to train their own AI models for image generation, which is central to the video's theme of personalizing AI-generated content.

💡Model training

Model training is the process of teaching a machine learning model to recognize and replicate specific patterns or features. In the video, it is used to train the AI to generate images with a particular celebrity's face, such as Selena Gomez.

💡Data set

A data set is a collection of data used for analysis or machine learning. In the video, the user creates a data set by uploading photos of a specific celebrity to train the AI model.

💡Trigger words

Trigger words are specific phrases or terms that activate a certain function or response in a system. In the context of the video, 'Selena' is set as a trigger word to generate images featuring Selena Gomez.

💡Advanced config

Advanced config refers to the more complex settings that allow users to fine-tune their model's performance. The video suggests that while beginners can use default settings, advanced users can adjust these configurations for better control over the AI model's behavior.

💡Start training

This is the action taken by the user to initiate the model training process. In the video, the user clicks 'start training' after setting up their data set and configurations, which begins the process of teaching the AI to generate specific images.

💡Base model

A base model is a pre-existing AI model that serves as a foundation for further training. The video mentions 'dream shaper' as an example of a base model that can be used to start the training process.

💡Training notification

A training notification is an alert received once the AI model has finished training. It signifies that the model is ready to be used, as mentioned in the video when the user receives a notification after the training process is complete.


In the context of the video, 'generate' refers to the action of creating an AI-generated image using the trained model. The user types in a prompt, including the model's trigger word, to produce a personalized image.

💡Dream Shaper model

The Dream Shaper model is a specific type of AI model mentioned in the video that can be used for generating images. It is one of the options provided by the Sea Art platform for users to choose from when training their models.


Explore the world of AI image generation using personal or favorite celebrity photos.

Utilize the Sea Art platform's model training feature for custom image creation.

Create a dataset with a name and category on the Sea Art website.

Edit the dataset by uploading personal images or those of a chosen celebrity.

Use the tag settings to set a model trigger word, such as a celebrity's name.

The platform suggests keywords for your images after uploading.

Save the tags after setting them to proceed with the model training.

Adjust the advanced configuration for fine-tuning, if desired.

Begin the training process by providing a unique name and selecting a base model.

Training may take some time, so relax while the process completes.

Receive a notification when the model is trained and ready to use.

Navigate to the LoRA tab to select and use the newly trained model.

Generate images by typing in a prompt that includes the model's trigger word.

Witness the creation of the first AI-generated image featuring the chosen celebrity.

Experiment with different models like the Dream Shaper for varied results.

Achieve seamless AI-generated images with the Sea Art training feature.

Share your creations and experiences with the community.

Join the Sea Art platform to start training your own AI model today.